In recent years, magnetic resonance imaging (MRI) has emerged as a powerful tool in clinical settings because it is noninvasive and yields an accurate volume rendering of the subject. The image is created by imposing one or more orthogonal magnetic field gradients upon the specimen while exciting nuclear spins with radio frequency pulses as in a typical nuclear magnetic resonance (NMR) experiment. After collection of data with a variety of gradient fields, deconvolution yields a one, two, or three dimensional image of the specimen. Typically, the image is based upon the NMR signal from the protons of water where the signal intensity in a given volume element is a function of the water concentration and relaxation times (T.sub.1 and T.sub.2). Local variations in these three parameters provide the vivid contrast observed in MR images. For example, the low water content of bone makes it distinctively dark, while the short T.sub.2 of clotted blood affords it a higher signal intensity than that from non-clotted blood.
The same advantages that have made MRI the technique of choice in medical imaging make it an ideal imaging tool for use in biological experiments. Unlike light-microscope imaging techniques based upon the use of dyes or fluorochromes, MRI does not produce toxic photobleaching by-products. Furthermore, unlike light-microscopy, MRI is not limited by light scattering or other optical aberrations to cells within approximately only one hundred microns of the surface. MRI was originally considered a purely noninvasive approach but more recently it has been found that contrast agents can significantly improve the diagnostic utility of the technique. MRI contrast agents dramatically reduce the relaxation times of protons in the surrounding water. The ion Gd.sup.3+, in its non-toxic chelated forms, is the most commonly used paramagnetic ion because of its large magnetic dipole and large effect on relaxation times. For example, Gd.sup.3+ chelated with diethylenetriaminepentaacetic acid (DTPA) is a vascular contrast agent now widely used in diagnostic radiology. The chemical structure of DTPA is depicted in FIG. 4.
Traditional MRI offers high spatial resolution and multiple plane imaging in a fast noninvasive procedure. When MRI contrast agents are used diagnostically, they are vascularly perfused, enhancing the contrast of blood vessels and reporting on organ lesions and infiltration. However, the labeling of specific tissues for diagnostic radiology remains a difficult challenge for MRI. Efforts to develop cell and tissue-specific MRI contrast agents by modifying existing immunological techniques has been the focus of much research in diagnostic radiology. For example, antibodies labeled with paramagnetic ions, generally the gadolinium chelate Gd-DTPA, have been generated and tested for their effects on MRI contrast of tumors and other tissues [Lex, Acta Biochim. Biophys. Hung. 24:265-281 (1989); U.S. Pat. No. 5,059,415]. It was anticipated that due to reductions in the rate of molecular tumbling, Gd-DTPA when bound to antibodies would show significantly higher relaxivity, a measure of MRI contrast enhancement, than that of unbound Gd-DTPA. This increase in relaxivity per Gd ion, it was hoped, would generate sufficient signal for tissue contrast to be observed using antibodies labeled with 10-50 Gd ions per protein molecule.
Unfortunately, the relaxivity of Gd bound to antibodies has been found to be only slightly better than that of unbound Gd-DTPA [Paajanen et al., Magn. Reson. Med 13:38-43 (1990)]. Therefore, to generate detectable contrast enhancement in an antibody-labeled tissue, the immunological reagent must be conjugated with hundreds if not thousands of Gd ions per antibody. Currently this is unattainable using standard techniques.
Several researchers have examined the possibility that the number of Gd ions per antibody could be increased by conjugating polylysine to the antibody, then labeling the polylysine extensively with Gd-DTPA [WO93/01837]. So far, these attempts have shown only limited success in part due to the unfavorable ionic and steric effects of conjugating antibodies to large polymers.
Research in the field of targeted MRI contrast agents has thus turned to the use of iron oxide particles as high signal strength T.sub.2 contrast agents [Shen et al., Magnet. Res. Med. 29:599-604 (1993); Weissleder et al., Magnetic Resonance Quarterly, 8:55-63 (1992)]. However, no iron oxide particles have yet been approved for use in humans.
Liposomes as carriers of contrast media show promise as tissue-specific MRI agents as well [Schwendener, R. A., Chimia 46:69-77 (1992)]. Two classes of such contrast agents have been developed: (i) water soluble contrast agents entrapped between phospholipid bilayers, and (ii) liposomes directly incorporating amphipatic molecules covalently attached to MRI contrast agents such as Gd-DTPA. The former class of liposomal contrast agents suffers from leakiness of the water soluble agent in vivo, and the later from long-term retention of the agent in the liver and spleen. Nevertheless, liposomes show promise as liver, spleen and lung contrast agents.